[PATCH 0/7] mm: Improve swap path scalability with batched operations

From: Tim Chen
Date: Tue May 03 2016 - 17:00:45 EST

The page swap out path is not scalable due to the numerous locks
acquired and released along the way, which are all executed on a page
by page basis, e.g.:

1. The acquisition of the mapping tree lock in swap cache when adding
a page to swap cache, and then again when deleting a page from swap cache after
it has been swapped out.Â
2. The acquisition of the lock on swap device to allocate a swap slot for
a page to be swapped out.Â

With the advent of high speed block devices that's several ordersÂÂ
of magnitude faster than the old spinning disks, these bottlenecks
become fairly significant, especially on server class machines
with many theads running.ÂÂTo reduce these locking costs, this patch
series attempt to batch the pages on the following oprations needed
on for swap:
1. Allocate swap slots in large batches, so locks on the swap device
don't need to be acquired as often.Â
2. Add anonymous pages to the swap cache for the same swap device inÂÂÂÂÂÂÂÂÂÂÂÂÂ
batches, so the mapping tree lock can be acquired less.
3. Delete pages from swap cache also in batches.

We experimented the effect of this patches. We set up N threads to access
memory in excess of memory capcity, causing swap.ÂÂIn experiments using
a single pmem based fast block device on a 2 socket machine, we saw
that for 1 thread, there is a ~25% increase in swap throughput and for
16 threads, the swap throughput increase by ~85%, when compared with the
vanilla kernel. Batching helps even for 1 thread because of contention
with kswapd when doing direct memory reclaim.

Feedbacks and reviews to this patch series are much appreciated.



Tim Chen (7):
 mm: Cleanup - Reorganize the shrink_page_list code into smaller
 mm: Group the processing of anonymous pages to be swapped in
 mm: Add new functions to allocate swap slots in batches
 mm: Shrink page list batch allocates swap slots for page swapping
 mm: Batch addtion of pages to swap cache
 mm: Cleanup - Reorganize code to group handling of page
 mm: Batch unmapping of pages that are in swap cache

Âinclude/linux/swap.h |ÂÂ29 ++-
Âmm/swap_state.cÂÂÂÂÂÂ| 253 +++++++++++++-----
Âmm/swapfile.cÂÂÂÂÂÂÂÂ| 215 +++++++++++++--
Âmm/vmscan.cÂÂÂÂÂÂÂÂÂÂ| 725 ++++++++++++++++++++++++++++++++++++++-------------
Â4 files changed, 945 insertions(+), 277 deletions(-)